Magnetic resonance brain perfusion imaging with voxel-specific arterial input functions

Area of Science:

• Neuroimaging
• Medical Physics
• Biomedical Engineering

Background:

• Dynamic contrast-enhanced (DCE) brain perfusion imaging is crucial for assessing cerebral hemodynamics.
• Accurate estimation of the arterial input function (AIF) is essential for quantitative analysis in DCE imaging.
• Manual AIF selection is time-consuming and prone to inter-observer variability.

Purpose of the Study:

• To develop and validate an automated method for estimating voxel-specific arterial input functions (AIFs) in dynamic contrast brain perfusion imaging.
• To overcome the limitations of manual AIF selection in terms of time and reproducibility.
• To improve the differentiation of hemodynamic parameters between different brain tissues.

Main Methods:

• Voxel-specific AIFs were estimated using homomorphic transformations and complex cepstrum analysis.
• Wiener filtering was applied for deconvolution of the estimated AIFs.
• The method was validated with simulated data and tested in 10 healthy adult subjects.

Main Results:

• Simulations demonstrated accurate estimation of diverse, normalized AIFs.
• While simple Wiener filtering led to flow underestimation, the new method showed comparable gray matter (GM) to white matter (WM) flow ratios to manual AIF selection.
• The automated method revealed significant differences in mean transit time (MTT) and time-to-peak (TTP) between GM and WM.

Conclusions:

• The proposed automated method for AIF estimation shows promise in replacing manual selection.
• This technique offers improved differentiation of time-dependent hemodynamic parameters, particularly MTT and TTP, between GM and WM.
• The findings suggest potential for enhanced quantitative analysis in brain perfusion imaging.